Available online at www.worldscientificnews.com
WSN 95 (2018) 75-88
EISSN 2392-2192
Yield Stability Analysis of Open Pollinated Maize
(Zea mays L.) and their Topcross Hybrids in Uganda
Netsanet Abera Muluneh1,2 *, Thomas Lapaka Odong1, Lwanga Charles Kasozi3,
Richard Edema1, Paul Gibson1, Daniel Koime3
1
College of Agricultural and Environmental Sciences, Makerere University,
P.O.Box 7062, Kampala Uganda
2
Pawe Agricultural Research Center, Ethiopian Institute of Agricultural Research,
P.O.Box 25, Addis Ababa, Ethiopia
3
National Crops Resources Research Institute,
P.O.Box 7084, Namulonge, Kampala, Uganda
*E-mail address: nabera2004@gmail.com
ABSTRACT
The study was aimed at determining yield stability and adaptability patterns of a set of 65 open
pollinated maize genotypes evaluated across four different agro ecologies in Uganda using 5 × 13 αlattice design replicated twice. Individual location analysis ANOVA results showed mean squares of
genotype were statistically highly significant in terms of days to 50% anthesis, anthesis silking
interval, grain yield and maize streak virus disease severity score for all environments tested except for
grain yield in Ngetta. The highest grain yield was recorded for topcross C9/TA (ECAVL1/CML536)
of 9.60 t ha-1 in Bulindi, for top cross C3/TA (Longe5/CML536) of 9.56 t ha-1 in Namulonge.
However, they were quite unstable as their ranking was not consistent across environments. The
genotype Ambsyn5, C4/TB, FS85 and C9/TB were showed the lowest disease score for MSV. The
parent OPV SITUKA MI was with the lowest day requirement for shading pollen and hence it can be
utilized in breeding for earliness. The additive main effects and multiplicative interaction (AMMI)
analysis results indicated that the tested genotypes were highly influenced by genotype main effects,
environment effects and genotype x environment interaction effects; the magnitude of environment
and its interaction effect for grain yield was 9.8 times greater than the variation attributed to genotype
main effects thus these genotypes were more affected by the environment and their interaction. Based
( Received 15 February 2018; Accepted 01 March 2018; Date of Publication 02 March 2018 )
World Scientific News 95 (2018) 75-88
on Finlay and Wilkinson’s sensitivity estimate, genotypes G40, G58, G42, G44, G56, G23, G52 and
G53 were identified as the most stable and widely adaptable.
Keywords: AMMI, G × E, Open pollinated varieties, Topcross, Yield stability, Zea mays
1. INTRODUCTION
Maize (Zea mays L.) is the principal cereal crop in Uganda. It remains the most
important food security crop in eastern and southern Africa (ESA) predominantly grown by
the resource-constrained and small-scale farmers. Newly introduced and local OPVs including
their newly generated topcrosses should exhibit great yield potential and average stability
over a wide range of agro-ecologies in Uganda. Stability of yield is defined as the ability of a
genotype to avoid substantial fluctuations in yield over a range of environment [1]. Cultivar
performance is a function of the genotype and the nature of the production environment [2].
In most cases maize productivity is function of genotype, environment and the genotype ×
environment interaction [3,4]. The differential response of a genotype across environments is
defined as the genotype (G) × environment (E) interaction. [6,7] indicated that it is the rule to
perform G × E in most quantitative characteristics. Grain yield in nature, routinely exhibits G
× E interaction [8] which necessitates evaluation of cultivars in multiple environments [9,10].
Crop cultivars are grown in diverse environments of different soil types, soil fertility levels,
moisture levels, temperatures and cultural practices. During production, all these cumulated
conditions constitute the growing environment for the crop varieties [11,25-28].
Existence of G × E interaction necessitates that breeders evaluate genotypes in more
than one environment to obtain repeatable rankings of genotypes. In maize breeding, choice
of a suitable candidate cultivar is subject to two considerations: (1) high grain yield across a
wide range of environments and (2) consistent performance over environments. The
performance of a genotype is determined by three factors: genotypic main effect (G),
environmental main effect (E) and their interaction (G × E) [12]. Consistency of performance
is dependent upon G × E interaction. Major difference in genotypes stability is due to
crossover interaction effect of genotype and environment. Genotype x Environment (G × E)
interaction study is an important common phenomenon in maize, especially when yield
stability of varieties is going to be studied [13,14]. This study is therefore mainly intended to
determine the yield stability (G × E) of the parental OPVs and their topcross hybrids in
Uganda.
2. MATERIAL AND METHOD
The research was conducted at four sites: i) The National Crops Resources Research
Institute (NaCRRI) Namulonge (Central Uganda) ii), National Semi-Arid Resources Research
Institute (NaSARRI) Serere (Eastern Uganda) iii) Bulindi Zonal Agricultural and
Development Research Institute (Western Uganda) and iv) Ngetta Zonal Agricultural and
Development Research Institute, Ngetta (Northern Uganda). The description of the study sites
for this study was conducted in 3 optimum sites and one site that expected random drought.
-76-
World Scientific News 95 (2018) 75-88
2. 1. Genetic material
The genetic materials in this section were 65 entries consisting of 19 OPVs, 38
topcrosses generated in 2015A and 8 checks. Additionally, these materials were considered to
be different in their genetic back ground from where they are sourced.
2. 2. Experimental design
All 65 entries were planted under rain-fed condition and evaluated in four different
agro-ecologies of Uganda. The experimental design used was 5 x 13 α-Lattice design with
two replications per location. Two row plots of 5 m long were used with an inter-row spacing
of 0.75 m and intra-row spacing of 0.25 m.
2. 3. Data collection
Key agronomic traits were measured for entire genotypes. Grain yield (t ha-1) and other
selected traits were the major trait of interest in this study. Data collected at vegetative and
flowering stage (Before harvest) were: Days to anthesis (AD), Anthesis-silking interval (ASI),
Maize strike virus (MSV), and at harvest: Grain yield (GY).
…………………………….........…..(1)
where: GY = grain yield, t= ton, ha-1 = per hectare, F.W. = Fresh weight of ear in kg at
harvest, MC = Grain moisture content at harvest, S = Shelling co-efficient (0.80), moisture
content required in grain at storage 12.5%, 1hectare = 10,000 m2, net area/plot = 7.5 m2 with 5
m long, 75 cm wide and 2 row plot.
3. STATISTICAL ANALYSIS
Analysis of variance was used to analyse differences between the 65 entries. The final
grain yield was subjected to analysis of variance (ANOVA) for each site and for all the sites
combined. For the combined analysis, genotype effect was assumed to be fixed and location
effects as random. The linear model used for individual site analysis was:
Y
μ +G +R +𝐵
+ε
…………………………………………….………. (2)
where: Yijk = the observed value of trait from the ith genotype from the kth block nested in the
jth replicate, µ is the overall mean, Gi = the effect of ith genotype, R is the effect of the jth
replication and 𝐵
is the effect kth block nested within the jth replicate and ε is a random
error term (ε ~𝑁 0, 𝜎 ).
The combined analysis of variance across locations was done using ANOVA in GenStat
and AMMI model was used to measure genotype stability. The Linear model used for
combined analysis was:
Y
μ +G +E + 𝐸 +ε
…………………………………………….……..….…. (3)
-77-
World Scientific News 95 (2018) 75-88
where: µ is the overall mean; Gi, Ej and GEij represent the effect of the genotype,
environment, and genotype-environment interaction, respectively; and ε is the average of the
random errors associated with the lth plot that receives the ith genotype in the jth environment.
Finlay-Wilkinson Regression (joint regression) analysis method [15] was used for
genotype stability analysis. The Linear model used for yield stability analysis was:
μ +
+𝐸
+
+
……………………………………...(4)
where:
is mean of individual genotype crosses, µ is the overall mean, is the effect of the
th
i genotype (fixed effects of genotype),
is measures the sensitivity of genotype i over the
sampled environments, 𝐸 is the effect of jth environment (random effect) and
is the
residual (lack of fit).
4. RESULTS
4. 1. Individual location analysis
The mean squares of each trait are shown in (Table 1). Results of ANOVA for each
individual location revealed statistical significant differences at (P < 0.05) for most traits
except for grain yield in Ngetta.
Mean of selected genotypes (from top four and bottom four) for traits days to anthesis,
anthesis silking interval, grain yield and maize streak virus in each location is indicated in
-78-
World Scientific News 95 (2018) 75-88
(Table 2 and Table 3) respectively. In terms of days to anthesis, the highest was recorded for
OPV Ambsyn2 (85.7 days) in Namulonge, 77 days in Bulindi, 76.5 days in Serere and 72.5
days in Ngetta while the lowest was for SITUK MI (54, 56.5, 57.5 and 62.9 days) in Ngetta,
Serere, Bulindi and Namulonge respectively. The highest anthesis silking interval of 4 days
was recorded for topcross C14/TA (KakSyn-II/CML536) in Ngetta, 3 days for varietal hybrid
check (UH5053) in Bulindi, 2.9 days for Ambsyn2 in Serere and 2 days for SUWAN in
Namulonge whereas the lowest was recorded -1.5 days for topcross C1/TA (MM3/CML536)
in Namulonge, 0.1 day for topcross C18/TB (SUWAN/CML202) in Serere, 0.5 days for
topcross C5/TA (Longe5D/CML536) in Bulindi and 1.5 days for topcross C19/TA (VP
MAX/CML536) in Ngetta.
In terms of yield, the highest was recorded for topcross C9/TA (ECAV1/CML536) of
9.60 t ha-1 in Bulindi, for topcross C3/TA (Longe5/CML536) of 9.6 t ha-1 in Namulonge, for
UH5354 (TWC) hybrid check with 4.77 t ha-1 in Serere and for topcross C9/TB
(TMV1/CML202) with 3.47 t ha-1 in Ngetta while the lowest was recorded for OPV Ambsyn2
with 0.95 t ha-1, 1 t ha-1, 1.46 t ha-1 in Serere, Ngetta and Namulonge respectively. The
topcross C15/TB (Ambsyn2/CML202) of 3.40 t ha-1 had the lowest grain yield recorded in
Bulindi. Regarding the maize streak virus diseases severity score, the highest severity was
recorded for topcrosses C1/TA (MM3/CML536) with value 3.56 in Bulindi, for topcross
C11/TA (ECAVL2/CML536) with value 3.09 in Namulonge, for topcross C14/TA (KakSynII/CML536) with value 2.26 in Serere, and for UH5053 (varietal cross check) with value 2.25
in Ngetta whereas the lowest was for topcross C4/TB (Longe5/CML202) with value 0.91 in
-79-
World Scientific News 95 (2018) 75-88
Serere, for Ambsyn5 with value 1.12 in Namulonge, for topcross C9/TB (TMV1/CML202)
with value 1.49 in Ngetta and FS85 with value 1.51 Bulindi see Table 3.
Table 3. Mean of top six and bottom five selected genotypes for grain yield and maize streak
virus in each of 4 locations in Uganda, 2015B
Grain yield (t ha-1)
Genotype
Namulonge
Serere
Bulindi
Ngetta
Ambsyn2
1.46
0.95
5.63
1.01
C9/TA
6.95
3.49
9.60
2.24
UH5354
6.97
4.77
4.41
1.96
C9/TB
7.10
3.43
6.17
3.47
C5/TA
9.23
3.49
6.79
1.84
C6/TA
9.28
4.67
6.45
2.54
C12/TA
9.42
3.52
7.71
1.73
C16/TA
9.48
3.62
8.70
2.14
C3/TA
9.56
3.50
6.74
1.94
Mean
6.86
3.27
6.14
1.96
% CV
15.18
22.35
23.02
29.14
LSD (5 %)
2.12
1.72
1.53
1.41
Table 3(continue). Mean of top six and bottom five selected genotypes for grain yield and
maize streak virus in each of 4 locations in Uganda, 2015B
Genotype
Maize streak virus score (1-5)
Namulonge
Serere
Bulindi
Ngetta
Ambsyn5
1.12
1.39
2.52
1.51
Longe 4
1.14
1.13
2.25
1.75
FS85
1.15
1.11
1.51
1.75
C1/TA
1.41
1.47
3.56
1.5
C9/TB
1.62
1.49
2.7
1.49
-80-
World Scientific News 95 (2018) 75-88
UH5053
1.96
1.36
1.8
2.25
C14/TA
2.48
2.26
3.53
1.75
C4/TB
2.58
0.91
1.72
2
C11/TA
3.09
1.55
2.76
1.51
Mean
1.94
1.34
2.52
1.80
% CV
27.00
20.06
23.3
11.10
LSD (5 %)
1.15
0.56
0.58
0.38
% cv-coefficient of variation, LSD- Least significant difference at 5 %, Min- minimum value,
Max-Maximum value.
4. 2. Combined analysis of variance for G x E
The combined analysis of variance (ANOVA) of the 65 entries evaluated across 4
locations according to the Additive Main effects and Multiplicative Interaction (AMMI)
model is presented in (Table 4). AMMI analysis indicated highly significant differences (P <
0.001) for environments (E), genotypes (G) and GxE interaction effects for all selected traits
except which were non-significant for anthesis silking interval and maize streak virus. The
first two interaction principle components (IPCA1 and IPCA2) were also highly significant (P
< 0.001).
The highest significant percentage of the total explained variation (51.44%) due to
environment effect was observed for grain yield. The first two IPCA axes explained 71.65 %
of the GxE interaction for days to anthesis, 74.90 % for anthesis silking interval, 91.57 % for
grain yield and 89.48 % for maize streak virus.
The first four AMMI selections per environment is summarized in Table 5. The first
two genotypes (G14 and G52) required the highest days to shade pollen in all environments.
The genotypes G6, G16, G18 and G30 were well performed genotypes in terms of yield in
more than two environments. The highest mean yield was observed for G3, G6, G12 and G16
in Namulonge and G9, G10 and G35 in Bulindi. Similar results were also observed when G x
E was analyzed using GGE biplot (result not shown).
The genotype sensitivity estimates for days to anthesis, anthesis silking interval, grain
yield and MSV was shown in Table 6. G28 and G65 were observed with the highest and the
lowest sensitivity estimate (0.89 and 0.49) for days to anthesis, G12 and G24 with sensitivity
(0.29 and 0.75) for anthesis silking interval, genotype 45 and 53 with (0.32 and 0.85) and
genotype 51 and 18 with (0.75 and –0.09) respectively.
5. DISCUSSION
5. 1. Individual location analysis
Determination of yield stability of genotypes evaluated in different agro ecologies is an
important goal for breeder to facilitate maize breeding and yield improvement efforts. For
-81-
World Scientific News 95 (2018) 75-88
most traits measured, except for grain yield in Ngetta indicated that there exists genetic
potential difference between the tested genotypes in each environment. These differences can
help the breeder to decide his breeding strategy to the specific environment for better yield.
Statistical analysis revealed significant differences among the tested maize genotypes
for days to anthesis. The OPV Ambsyn2 took maximum number of days for pollen shedding
(85.7 days) in Namulonge, 77 days in Bulindi, 76.5 days in Serere and 72.5 days in Ngetta
while the minimum date was for SITUK MI (54, 56.5, 57.5 and 62.9 days) in Ngetta, Serere,
Bulindi and Namulonge respectively indicated the existence of genotype by environment
interaction effect. Therefore, OPV Ambsyn2 and SITUKA MI were grouped under the
highest and lowest days to anthesis respectively. These contrasting ranges of days for
flowering can help breeders to breed these varieties in different maturity groups for tested
environments. The newly introduced genotype SITUKA MI noted to be quite the earliest
OPV among the genotypes evaluated and could be used as germplasm source in developing
varieties with early maturity. Previously, [16,17] have also reported significant amount of
variability for days to anthesis among different open pollinated maize genotypes.
Means of top eleven genotype ranking for anthesis-silking interval shown in (Table 4)
reveal significant amount of variation for ASI in each locations. The means of anthesis silking
interval that exhibited among the topcrosses (C8/TA, C17/TA, C19/TA) in Namulonge were
having 0.5 days i.e. both the pollen shedding and silking were well synchronized. This
synchronization helps to carry out successful pollination so as to produce better yield. At the
same time, the Genotypes (C1/TA, C17/TB, C10/TB, C4/TA, FS85, ECAVL1, ECAVL18)
gave negative ASI values ranging from -1.5 to -1 days in Namulonge. The observed negative
sign of ASI shows that the silking was earlier than pollen shedding (female flowered first)
indicating that there exists the possibility to develop improved drought tolerant open
pollinated maize varieties with better synchronization in anthesis and silking parameters while
positive sign of ASI indicates that pollen shedding was earlier than silking. Similarly,
genotypes C18/TB with (0.1 days) and TMV (0.5 days) in Serere, C5/TA, C15/TA, VP MAX
and C13/TA with (0.5 days) in Bulindi and C19/TA (1.5 days) in Ngetta were with the lowest
value of ASI. Therefore, these genotypes showed potential to breed for specific environment
with better pollen shading and silking time synchronizations. [18] observed considerable
genotypic variability among various maize genotypes for different traits.
Genotype ranking using mean grain yield is shown in (Table 5). Highly significant (p <
0.001) variability in grain yield among the genotypes evaluated in 4 location indicated
difference in responses for each genotype in the different environment. The highest grain
yield was recorded for topcross C9/TA (TMV1/CML536) of 9.60 t ha -1 in Bulindi and for
topcross C3/TA (Longe4/CML536) of 9.56 t ha-1 in Namulonge. Similarly, the topcrosses:
C16/TA (9.48 t ha-1), C12/TA (9.42 t ha-1), C6/TA (9.28 t ha-1) and C5/TA (9.23 t ha-1)
selected as high yielders. However, they were quite unstable as their ranking was not
consistent across environments. As it was observed from the result, better responses were
noted for topcrosses than parental OPVs and the checks indicating future possibilities to breed
for specific environment with higher yield. Souza et al., (2008) reported that different
genotypes have different performance in each region that can be capitalized to maximize
productivity.
Regarding the MSV diseases severity on a scale (1-5), the highest severity was recorded
for topcross C1/TA (MM3/CML536) with value 3.56 in Bulindi, for topcross C11/TA
(ECAVL2/CML536) with value 3.09 in Namulonge, for topcross C14/TA (KakSyn-
-82-
World Scientific News 95 (2018) 75-88
II/CML536) with value 2.26 in Serere, and for UH5053 (varietal cross check) with value 2.25
in Ngetta whereas the lowest was for topcross C4/TB (Longe5/CML202) with value 0.91 in
Serere, for Ambsyn5 with value 1.12 in Namulonge, for topcross C9/TB (TMV1/CML202)
with value 1.49 in Ngetta and FS85 with value 1.51 Bulindi indicating that the presence of
the promising genotypes resistance to MSV. [20] on six foliar disease based on severity score
has explained the variation among tested genotypes which can help researchers to develop
disease resistant varieties.
Table 4. ANOVA results based on the AMMI model of days to anthesis, anthesis silking
interval, grain yield and maize streak virus for maize genotypes tested in four locations
showing the total sum of squares and the portion of the total sum of squares explained by
G, E, IPC1 and IPC2
Total variation explained by each component (%)
Trait
TSS
Genotype
Environment
IPCA 1
IPCA2
AD
11742.0
39.32 ***
51.44 ***
3.22 ***
2.25 ***
ASI
556.3
10.57 ns
53.33 ***
11.94 ***
9.04 ***
GY
3238.0
8.26 ***
68.01 ***
6.45 ***
5.02 ***
MSV
228.6
10.71 ns
39.07 ***
14.70 ***
10.08 ***
*, **, ***Statistically significant
00 ,00
0 00 respectively, ns - statistically non-significant,
E - Environment; G × E - genotype by Environment; Rep - replication; df - degrees of freedom; SS- Sum of
square; MS - Mean square; IPCA - Interaction principal component axis, TSS - total sum of squares
Table 5. The first four AMMI selections per environment for 65 tested maize OPV
genotypes in Uganda, 2015B.
Days to anthesis (days)
Anthesis silking interval (days)
Environment
Environment
Mean
1st
Ngetta
64.1
G52 G14 G45 G65
Bulindi
69.7
Serere
Namulonge
2nd
3rd
4th
Mean
1st
Namulonge
0.8
G39 G55 G63 G56
G52 G14 G45 G65
Ngetta
2.9
G52 G14 G63
G2
65.1
G52 G14 G45 G65
Serere
1.5
G52 G14
G10
72.5
G52 G14 G11 G64
Bulindi
1.5
G3
Grain yield (days)
3rd
G3
4th
G62 G21 G57
Maize streak virus (score 1-5)
Environment
Namulonge
2nd
Environment
Mean
1st
2nd
3rd
4th
6.97
G12 G16
G6
G3
-83-
Bulindi
Mean
1st
2nd
3rd
4th
2.5
G14
G1
G10 G64
World Scientific News 95 (2018) 75-88
Serere
3.26
G18 G30
G6
G19
Ngetta
1.8
G14
G7
G57 G32
Ngetta
1.91
G18 G28 G30 G11
Serere
1.3
G14
G4
G32
Bulindi
6.11
G9
Namulonge
1.9
G11 G10 G22 G14
G35 G16 G10
G9
G - genotype; AMMI - interaction additive main - effect and multiplicative interaction
Table 6. Finlay and Wilkinson modified joint regression analysis sorted sensitivity estimates
eight least sensitive (most stable) genotypes.
AD (days)
ASI (days)
G
β
SE
Mean
MS dev.
G
β
SE
Mean
MS dev.
65
0.49
0.21
71.84
5.69
24
0.28
0.37
1.53
0.09
34
0.54
0.21
67.40
2.46
21
0.38
0.37
1.76
0.46
45
0.56
0.21
72.86
2.06
19
0.38
0.37
1.26
0.21
51
0.57
0.21
67.69
1.39
51
0.40
0.37
1.70
0.15
8
0.64
0.21
70.10
6.22
54
0.45
0.37
1.46
0.30
14
0.64
0.21
73.81
0.14
7
0.46
0.37
1.53
0.01
49
0.88
0.21
70.19
0.72
40
0.74
0.37
1.50
0.07
28
0.89
0.21
66.95
0.12
12
0.75
0.37
1.51
0.09
GY (t ha-1)
MSV (score 1-5)
G
β
SE
Mean
MS dev.
G
β
SE
Mean
MS dev.
53
0.31
0.23
2.78
5.15
18
–0.08
0.46
1.88
0.01
52
0.48
0.23
2.26
5.61
35
–0.06
0.46
1.61
0.04
40
0.56
0.23
3.61
0.06
57
–0.02
0.46
1.79
0.02
58
0.56
0.23
3.70
0.88
33
0.23
0.46
1.73
0.11
42
0.57
0.23
3.63
1.86
62
0.25
0.46
1.84
0.18
44
0.60
0.23
3.57
1.09
58
0.26
0.46
1.38
0.11
56
0.63
0.23
4.16
0.93
4
0.29
0.46
2.16
0.18
45
0.85
0.23
3.76
0.68
51
0.75
0.46
1.86
0.06
AD - days to anthesis; ASI - anthesis silkig interval; G - genotype; β - genotype sensitivity estimate;
SE - standard error; MS dev. - mean square of deviation.
-84-
World Scientific News 95 (2018) 75-88
5. 2. Combined analysis of Variance for GxE by AMMI model
The interaction between genotypes and environment that produces a phenotype is
referred as genotype x environment interaction. AMMI analysis result for G × E showed
statistically highly significant effect of environments, genotypes and genotypes x environment
interaction on days to anthesis, anthesis-silking interval, grain yield and reaction to maize
streak virus disease. The observed performance of the tested maize genotypes indicated their
potential to respond and adapt the environment either broadly or specifically. The variation
due to environment and G × E was 1.5 time more than variation attributed to genotype main
effect for days to anthesis, 7.7 times more for anthesis silking interval, 9.8 times more for
grain yield and 6.2 time more for maize streak virus disease severity scores indicating the
importance of G × E interaction.
The occurrence of a statistically significant G × E interaction effect indicated
inconsistent phenotypic performance of the tested genotypes across locations, which may
cause selections made in one environment to perform poorly in another environment. In
agreement with this study, [21] reported the higher contribution of environmental variance
than both genotypic variance and G × E interaction in terms of grain yield for wheat cultivars
tested in 3 different locations. However, [22] in contrast reported that variation due to
genotype was higher than both the environment variance and G × E interaction for 30 maize
hybrids tested across locations in Kenya. The significant contribution of environment to the
differences in genotype expression across locations would remain a serious challenge in both
crop management and breeding for varieties with little variation (stable) or improved
performance due to environmental changes. Large contribution of the environment component
to grain production in maize has been similarly reported by many works [23,24]. Therefore,
differences in the amount of variation accounted for by the environment, the values obtained
were important implications to emphasize the necessity of description of environments for the
ultimate interest of driving the direction of plant breeding for stable or positive response.
Regarding the genotype stability, AMMI analysis result showed the occurrence of a
statistically significant G × E interaction effect. As a result, inconsistent performance of the
tested genotypes across locations was observed. However, according to [15] genotype
sensitivity estimate, the genotypes with the lowest sensitivity were the most stable and widely
adaptable. Therefore, in terms of days to anthesis (G65, G34, G45 and G51), anthesis silking
interval (G24, G21, G19 and G51), Grain yield (G40, G58, G42, G44, G56 G23 G52, G53)
and maize streak disease score (G18, G33, G35, G57, G62) were most stable and wider
adaptable.
6. CONCLUSIONS
Individual location analysis showed highly significant differences among the genotypes
evaluated. The topcrosses showed quite better response than the OPVs and the check in grain
yield. The results of AMMI analysis indicated that the genotype performance of maize
parental OPVs and topcrosses were highly influenced by genotype main effects, environment
effects and genotype x environment interaction effects; the magnitude of environment and its
interaction effect for grain yield was 9.8 times that of variation attributed due to genotype
main effects indicating that the importance of genotype by environment (the tested genotypes
were affected by G × E). Those genotypes with lower sensitivity estimate were most stable
-85-
World Scientific News 95 (2018) 75-88
and wide adapters they can give appropriate mean yield when grown in different agro
ecologies of Uganda. In general, high variability exists among the tested genotypes. The
presence of this phenotypic and genotypic variability among the local and introduce OPVs
indicates that the genotypes could be used in future maize breeding programs for further
manipulation to develop breeding lines with improved attributes. The promising topcrosses:
C3/TA, C5/TA, C6/TA, C9/TA, C12/TA, C9/TA, C9/TB, C11/TA, C15/TA and C16/TA due
to their superiority in grain yield and Ambsyn5, C4/TB, FS85 and C9/TB because of their
resistance to MSV can be incorporated in breeding program for specific locations. The parent
SITUKA MI due to its the lowest day requirement for shading pollen can be utilized in
breeding for earliness. Stable genotypes (i.e. wider adaptable across locations) in terms of
days to anthesis (G65, G34, G45 and G51), anthesis silking interval (G24, G21, G19 and
G51), grain yield (G40, G58, G42, G44, G56 G23 G52, G53) and maize streak disease score
(G18, G33, G35, G57, G62) can be included in the breeding programme targeting those traits.
In general genotypes with superior characters are recommended to be incorporated in maize
breeding program. However, this study was carried out only in 4 locations for one season and
due highly significant genotype by environment interaction for an effective stability
determination it is important to consider testing these genotypes in many locations and more
than two seasons.
Acknowledgement
The author is grateful to all authors for their professional help. Department of maize breeding program at
NaCRRI, Namulonge for their technical supports. The author is also grateful to AGRA (Alliance for Green
revolution in Africa), for the fellowship during both the course and research work.
Reference
[1]
M.S. Kang, Using Genotype by Environment Interaction for Crop Cultivar
Development. Adv. Agron. 62 (1998) 199–246.
[2]
M. Cooper, D.E. Byth, Understanding plant adaptation to achieve systematic applied
crop improvement- A fundamental challenge in Plant adaptation and crop improvement,
CAB Int. IRRI, UK. (1996) 5–23.
[3]
H. Namorato, G.V. Miranda, L.V. De Souza, L.R. Oliveira, R.O. Delima, Comparing
Biplot Multivariate Analyses with Eberhart and Russell ’ Method for Genotype x
Environment Interaction, Biotechnology 9 (2009) 299–307. doi:10.12702/19847033.v09n04a03.
[4]
A. Deitos, E. Arnhold, F. Mora, G.V. Miranda, Yield and Combining Ability of Maize
Cultivars under Different Ecogeographic Conditions. Crop Breed. Appl. Biotechnol. 6
(2006) 222–227.
[5]
Y. Beyene, S. Mugo, K. Pillay, T. Tefera, S. Ajanga, S. Njoka, H. Karaya, J. Gakunga,
Testcross Performance of Doubled Haploid Maize Lines Derived from Tropical
Adapted Backcross Populations, Maydica 56 (2011) 351–358.
-86-
World Scientific News 95 (2018) 75-88
[6]
Y. Beyene, S. Mugo, C. Mutinda, T. Tefera, H. Karaya, S. Ajanga, J. Shuma, R. Tende,
V. Kega, Genotype by Environment interactions and Yield Stability of Stem Borer
Resistant Maize Hybrids in Kenya, African J. Biotechnol. 10 (2011) 4752–4758.
[7]
J.M. Bernardo, R. Rueda, Bayesian hypothesis testing: A reference approach. Int. Stat.
Rev. 70 (2002) 351–372. doi:10.1111/j.1751-5823.2002.tb00175.x.
[8]
I.A. Khalil, H.U.R. Rahman, M. Arif, I.H. Khalil, M. Iqbal, K. Afridi, M. Sajjad, M.
Ishaq, Evaluation of Maize Hybrids for Grain Yield Stability in North-West of Pakistan.
Sarhad J. Agric. 27 (2011) 213–218.
[9]
X. Fan, M.S. Kang, H. Chen, Y. Zhang, J. Tan, C. Xu, Yield Stability of Maize Hybrids
Evaluated in Multi-Environment Trials in Yunnan, China. Am. Soc. Agron. 99 (2007)
220–228. doi:10.2134/agronj2006.0144.
[10] T. Bacha, Genotype X Environment Interaction and Yield Stabliity of Breade Wheat
(Triticum eastivum L .) Genotype in Ethiopia using the Ammi Analisis. J. Biol. Agric.
Healthc. 5 (2015) 129–140.
[11] M.S. Abdulai, P.Y.K. Sallah, O. Safo-Kantanka, Maize Grain Yield Stability Analysis
in Full Season Lowland Maize in Ghana. Int. J. Agric. Biol. 9 (2007) 41–45.
[12] W. Yan, M.S. Kang, B. Ma, S. Woods, P.L. Cornelius, GGE Biplot vs AMMI Analysis
of Genotype-by-Environment Data, Crop Sci. 47 (2007) 641–653.
doi:10.2135/cropsci2006.06.0374.
[13] W. Abera, V. Rensburg, M. Labuschagne, H. Maartens, Genotype-Environment
Interactions and Phenotypic Stability Analyses of Linseed in Ethiopia, S. Afri. J. Plant
Soil. 71 (2004) 66–71.
[14] M. Shahryarinasab, R. Chogan, Stability Analysis of Maize Yield by using Unvariate
Statistical Method, Biol. Forum – An Int. J. 7 (2015) 394–399.
[15] K.W. Finlay, G.N. Wilkinson, The Analysis of Adaptation in a Plant-Breeding
Programme, Aust. J. Agric. Res. 14 (1963) 742–754.
[16] H.- Rahman, I. Khalil, I. Hussain, F. Afzal, Diversity for Morphological and Maturity
Traits in Maize Populations From Upper Dir, 24 (2008).
[17] A.E. Idris, A.I. Abuali, Genetic variability for vegetative and yield traits in maize (Zea
mays L) genotypes, Int. Res. J. Agric. Sci. Soil Sci. 1 (2011) 408–411.
[18] F. Ali, I. Nawaz, M. Munir, H. Ullah, Genetic Analysis of Mmaturity and
Morphological Traits under Maydis Leaf Blight ( MLB ) Epiphytoticsin Maize ( Zea
mays L.), ARPN J. Agric. Biol. Sci. 6 (2011) 13–19.
[19] M.F.M. De Souza, R. Rodrigues, A.T. Do Amaral, C.P. Sudré, Resistance to
Xanthomonas spp. in Tomato: Diallel analysis and gene effects estimative in a breeding
programme carried out in Brazil, J. Phytopathol. 156 (2008) 660–667.
doi:10.1111/j.1439-0434.2008.01474.x.
[20] X. Wang, Y. Zhang, X. Xu, H. Li, X. Wu, S. Zhang, X. Li, Evaluation of maize inbred
lines currently used in Chinese breeding programs for resistance to six foliar diseases,
Crop J. 2 (2014) 213–222. doi:10.1016/j.cj.2014.04.004.
-87-
World Scientific News 95 (2018) 75-88
[21] K. Koutis, Multienvironmental evaluation of wheat landraces by GGE Biplot Analysis
for organic breeding, Agric. Sci. 03 (2012) 66–74. doi:10.4236/as.2012.31009.
[22] Y. Beyene, S. Mugo, T. Tefera, J. Gethi, J. Gakunga, H. Karaya, R. Musila, W. Muasya,
R. Tende, S. Njoka, Yield Stability of Stem Borer Resistant Maize Hybrids Evaluated in
Regional Trials in East Africa, African J. Plant Sci. 6 (2012) 77–83.
doi:10.5897/AJPS11.262.
[23] A.L. Nassir, O.J. Ariyo, Genotype x environment interaction and yield-stability
analyses of rice grown in tropical inland swamp, Not. Bot. Horti Agrobot. Cluj-Napoca.
39 (2011) 220–225.
[24] M. Kandus, D. Almorza, R.B. Ronceros, J. Salerno, Statistical Models for Evaluating
the Genotype-Environment Interaction in Maize (Zea mays L), Int. J. Exp. Bot. 79
(2010) 39–46.
[25] Duru Chidi Edbert, Duru Ijeoma Akunna, Adsorption Capacity of Maize Biomass Parts
in the Remediation of Cu2+ Ion Polluted Water. World News of Natural Sciences 12
(2017) 51-62
[26] Yallew Mazengia. Smallholders commercialization of maize production in Guangua
district, northwestern Ethiopia. World Scientific News 58 (2016) 65-83
[27] P. I. Akwaji, E. J. Umana, E. I. Okon. Phytochemical and Antifungal Activity of Leaf
Extracts of Corchorus olitorius and Gongronema latifolium on Fungi Associated with
Post-Harvest Deterioration of Maize (Zea mays) Seeds in Oban Community, Nigeria.
World Scientific News 53(3) (2016) 157-177
[28] Musefa Redi, Wubayehu Gebremedhin, Fitsum Merkeb, Mohamed Yimam. Critical
Level of Extractable Phosphorus for Maize (Zea mays L.) at Metekel Zone,
Northwestern Ethiopia. World Scientific News 54 (2016) 14-26
-88-